Automatically generating inventory-related information forecasts using machine learning techniques

Abstract

Methods, apparatus, and processor-readable storage media for automatically generating inventory-related information forecasts using machine learning techniques are provided herein. An example computer-implemented method includes training a machine learning model for calculating weights for multiple statistical forecasts for inventory-related data by processing historical data pertaining to at least a portion of the multiple statistical forecasts; generating two or more statistical forecasts for inventory-related data associated with an enterprise by processing data pertaining to multiple system parts across multiple geographic granularities associated with the enterprise; calculating weights for the generated statistical forecasts by processing data associated with at least a portion of the generated statistical forecasts using the trained machine learning model; generating at least one composite inventory-related forecast by combining the generated statistical forecasts in accordance with the calculated weights; and performing automated actions based on the at least one composite inventory-related forecast.

What is claimed is: 
     
       1. A computer-implemented method comprising:
 training at least one machine learning model for calculating weights for multiple statistical forecasts for inventory-related data by processing historical data pertaining to at least a portion of the multiple statistical forecasts; 
 generating two or more statistical forecasts for inventory-related data associated with at least one enterprise by processing data pertaining to multiple system parts across multiple geographic locations associated with the at least one enterprise, wherein processing data pertaining to multiple system parts across multiple geographic locations comprises processing one or more of data pertaining to demand history for one or more of the multiple system parts across one or more of the multiple geographic locations, and processing data pertaining to one or more of the multiple system parts associated with at least one predefined enterprise service identifier; 
 calculating weights for the two or more generated statistical forecasts, based at least in part on accuracy of at least one forecasting methodology associated with the two or more generated statistical forecasts, for the inventory-related data associated with the at least one enterprise by processing data associated with at least a portion of the two or more generated statistical forecasts using the at least one machine learning model; 
 generating at least one composite inventory-related forecast for the at least one enterprise by combining the two or more generated statistical forecasts in accordance with the calculated weights; and 
 performing one or more automated actions based at least in part on the at least one composite inventory-related forecast, wherein performing the one or more automated actions comprises:
 generating and outputting one or more inventory-related recommendations, based at least in part on the at least one composite inventory-related forecast, to one or more users associated with the at least one enterprise; 
 automatically retraining at least a portion of the at least one machine learning model based at least in part on the at least one composite inventory-related forecast; and 
 automatically initiating provisioning of one or more of the multiple system parts, in amounts related to the at least one composite inventory-related forecast, to one or more of the multiple geographic locations associated with the at least one enterprise; 
 
 wherein the method is performed by at least one processing device comprising a processor coupled to a memory. 
 
     
     
       2. The computer-implemented method of  claim 1 , wherein processing data pertaining to multiple system parts across multiple geographic locations comprises processing data pertaining to location share associated with at least a portion of the multiple system parts. 
     
     
       3. The computer-implemented method of  claim 1 , wherein processing data pertaining to multiple system parts across multiple geographic locations comprises processing data pertaining to demand at multiple system part-related levels and multiple geographic location-related levels. 
     
     
       4. The computer-implemented method of  claim 1 , wherein training the at least one machine learning model comprises processing the historical data pertaining to at least a portion of the multiple statistical forecasts using at least one stochastic gradient descent technique. 
     
     
       5. The computer-implemented method of  claim 4 , wherein training the at least one machine learning model comprises:
 generating at least one receiver operating characteristic curve. 
 
     
     
       6. The computer-implemented method of  claim 5 , wherein training the at least one machine learning model further comprises:
 processing at least a portion of data points in at least one area under the at least one receiver operating characteristic curve using the at least one stochastic gradient descent technique. 
 
     
     
       7. The computer-implemented method of  claim 1 , further comprising:
 removing one or more constraints from the at least one machine learning model using at least one Lagrangian function. 
 
     
     
       8. A non-transitory processor-readable storage medium having stored therein program code of one or more software programs, wherein the program code when executed by at least one processing device causes the at least one processing device:
 to train at least one machine learning model for calculating weights for multiple statistical forecasts for inventory-related data by processing historical data pertaining to at least a portion of the multiple statistical forecasts; 
 to generate two or more statistical forecasts for inventory-related data associated with at least one enterprise by processing data pertaining to multiple system parts across multiple geographic locations associated with the at least one enterprise, wherein processing data pertaining to multiple system parts across multiple geographic locations comprises processing one or more of data pertaining to demand history for one or more of the multiple system parts across one or more of the multiple geographic locations, and processing data pertaining to one or more of the multiple system parts associated with at least one predefined enterprise service identifier; 
 to calculate weights for the two or more generated statistical forecasts, based at least in part on accuracy of at least one forecasting methodology associated with the two or more generated statistical forecasts, for the inventory-related data associated with the at least one enterprise by processing data associated with at least a portion of the two or more generated statistical forecasts using the at least one machine learning model; 
 to generate at least one composite inventory-related forecast for the at least one enterprise by combining the two or more generated statistical forecasts in accordance with the calculated weights; and 
 to perform one or more automated actions based at least in part on the at least one composite inventory-related forecast, wherein performing the one or more automated actions comprises:
 generating and outputting one or more inventory-related recommendations, based at least in part on the at least one composite inventory-related forecast, to one or more users associated with the at least one enterprise; 
 automatically retraining at least a portion of the at least one machine learning model based at least in part on the at least one composite inventory-related forecast; and 
 automatically initiating provisioning of one or more of the multiple system parts, in amounts related to the at least one composite inventory-related forecast, to one or more of the multiple geographic locations associated with the at least one enterprise. 
 
 
     
     
       9. The non-transitory processor-readable storage medium of  claim 8 , wherein processing data pertaining to multiple system parts across multiple geographic locations comprises processing data pertaining to location share associated with at least a portion of the multiple system parts. 
     
     
       10. The non-transitory processor-readable storage medium of  claim 8 , wherein training the at least one machine learning model comprises processing the historical data pertaining to at least a portion of the multiple statistical forecasts using at least one stochastic gradient descent technique. 
     
     
       11. The non-transitory processor-readable storage medium of  claim 10 , wherein training the at least one machine learning model comprises:
 generating at least one receiver operating characteristic curve. 
 
     
     
       12. The non-transitory processor-readable storage medium of  claim 11 , wherein training the at least one machine learning model further comprises:
 processing at least a portion of data points in at least one area under the at least one receiver operating characteristic curve using the at least one stochastic gradient descent technique. 
 
     
     
       13. The non-transitory processor-readable storage medium of  claim 8 , wherein the program code when executed by the at least one processing device further causes the at least one processing device:
 to remove one or more constraints from the at least one machine learning model using at least one Lagrangian function. 
 
     
     
       14. An apparatus comprising:
 at least one processing device comprising a processor coupled to a memory; 
 the at least one processing device being configured:
 to train at least one machine learning model for calculating weights for multiple statistical forecasts for inventory-related data by processing historical data pertaining to at least a portion of the multiple statistical forecasts; 
 to generate two or more statistical forecasts for inventory-related data associated with at least one enterprise by processing data pertaining to multiple system parts across multiple geographic locations associated with the at least one enterprise, wherein processing data pertaining to multiple system parts across multiple geographic locations comprises processing one or more of data pertaining to demand history for one or more of the multiple system parts across one or more of the multiple geographic locations, and processing data pertaining to one or more of the multiple system parts associated with at least one predefined enterprise service identifier; 
 to calculate weights for the two or more generated statistical forecasts, based at least in part on accuracy of at least one forecasting methodology associated with the two or more generated statistical forecasts, for the inventory-related data associated with the at least one enterprise by processing data associated with at least a portion of the two or more generated statistical forecasts using the at least one machine learning model; 
 to generate at least one composite inventory-related forecast for the at least one enterprise by combining the two or more generated statistical forecasts in accordance with the calculated weights; and 
 to perform one or more automated actions based at least in part on the at least one composite inventory-related forecast, wherein performing the one or more automated actions comprises:
 generating and outputting one or more inventory-related recommendations, based at least in part on the at least one composite inventory-related forecast, to one or more users associated with the at least one enterprise; 
 automatically retraining at least a portion of the at least one machine learning model based at least in part on the at least one composite inventory-related forecast; and 
 automatically initiating provisioning of one or more of the multiple system parts, in amounts related to the at least one composite inventory-related forecast, to one or more of the multiple geographic locations associated with the at least one enterprise. 
 
 
 
     
     
       15. The apparatus of  claim 14 , wherein processing data pertaining to multiple system parts across multiple geographic locations comprises processing data pertaining to location share associated with at least a portion of the multiple system parts. 
     
     
       16. The apparatus of  claim 14 , wherein processing data pertaining to multiple system parts across multiple geographic locations comprises processing data pertaining to demand at multiple system part-related levels and multiple geographic location-related levels. 
     
     
       17. The apparatus of  claim 14 , wherein training the at least one machine learning model comprises processing the historical data pertaining to at least a portion of the multiple statistical forecasts using at least one stochastic gradient descent technique. 
     
     
       18. The apparatus of  claim 17 , wherein training the at least one machine learning model comprises:
 generating at least one receiver operating characteristic curve. 
 
     
     
       19. The apparatus of  claim 14 , wherein the at least one processing device is further configured:
 to remove one or more constraints from the at least one machine learning model using at least one Lagrangian function. 
 
     
     
       20. The apparatus of  claim 18 , wherein training the at least one machine learning model further comprises:
 processing at least a portion of data points in at least one area under the at least one receiver operating characteristic curve using the at least one stochastic gradient descent technique.